Determination of Boron in Plant Material An Ignition-Electrometric Titration Method L. v. WILCOX U. S . Department of Agriculture, Division of Irrigation Agriculture, Rubidoux Laboratory, Riverside, Calif. KITION by Kelley and Brown
(4)in 1928 of the
depends upon the fact that a 0.7 N calomel and a quinhydrone electrode come to a null point (reversal of polarity) a t approximately p H 7 .
R E Ctoxlc O G effect of boron, occurring naturally in irrigation waters, upon citrus trees led to a n investigation of this problem by the Division of Western Irrigation Agriculture of the United States Department of Agriculture. I n connection with this study a method (7) for the determination of boron in plant material was worked out. This method is satisfactory for the determination of relatively large quantities of boron but i t is not well adapted to the study of low concentrations-e. g., under 50 mg. of boron per kilogram of dry material-because of a rather large blank and the relatively small samples that can be analyzed. The confirmation by Warington (6), that boron is essential to plant growth, and the demonstration of boron deficiency in the field by Brandenburg (2) in Germany and later by JIcMurtrey (5) in the United States, led to a very large number of investigations of boron deficiency. An accurate method for the determination of boron in these low concentrations was greatly needed but was not available. T o fill this need, a study was made of the possibilities of adapting the electrometric titration method (1, 8) to the determination of boron in plant material, and as a result a satisfactory method has been developed and is presented herein. The accuracy and reproducibility of the electrometric titration of boric acid were discussed in a n earlier paper (8). Since that time, the procedure has been used in this laboratory for the determination of boron in approximately 10,000 samples, mostly natural waters.
Galvanometer. An enclosed lamp and scale type sensitive to 0.025 microampere per scale division. Quinhydrone electrode. A piece of platinum wire 7.5 em. (3 inches) in length, with suitable contact above the surface of the solution, is preferable to an electrode of platinum sealed through glass and connected with mercury, as minute cracks develop in the glass and cause erratic results. Calomel electrode, 0.7 N with respect to potassium chloride. In the determination of boron in natural waters ( 1 , 8 ) , a silversilver chloride electrode was suggested. This can be used in place of the 0.7 iV calomel electrode in the method here described, but it is necessary to adjust the chloride concentration to some predetermined value and to standardize the sodium hydroxide in the presence of this concentration of chloride. The calomel electrode is recommended. Motor stirrer. Switch, single-pole single-throw. The electrodes are connected through the witch to the galvanometer. A shunt t o protect the galvanometer is desirable but not essential. This circuit diagram was shown in the earlier paper (1). The second apparatus is a simple potentiometer and is illustrated in Figure l . In addition to the parts listed above, the following are required: resistance wire, 1500 ohms tapped at 60 ohms; a 1.5-volt dry cell: and a calomel electrode, 0.1 N with respect to potassium chloride. This is substituted for the 0.7 A; electrode described above. The third apparatus makes use of 1.5 V O L T S either a potentiometer or “pH meter” as indicating system. The following
Reagents Used Lead nitrate solution, 1 N ; sodium bicarbonate, powder; cal-
cium oxide, fine powder. Quinhydrone, reagent quality, free from heavy metals. Bromothymol blue indicator solution, 1 per cent. Hydrochloric acid, 6 N , 2 N , and 0.02 N . Sodium hydroxide, 0.5 N , carbonate-free. Sodium hydroxide, standard 0.0231 N , carbonate-free (1ml. is equivalent to 0.25 mg. of boron). Boric acid solution. Dissolve 0.5716 gram of reagent grade, dry H3BOs in carbon dioxidefree distilled water and dilute to 1 liter. One milliliter contains 0.1 mg. of boron. This solution is used in standardizing the 0.0231 N sodium hydroxide. Mannitol (mannite), neutral. The blank titration for 5 grams of mannitol should not exceed 0.1 ml. of the standard 0.0231 N sodium hydroxide.
I
-1
ALVANOMETE R
0.1 N CALOMEL
FIGURE 1. CIRCUITDIAGRAM OF ELECTROMETRIC TITRATION APPARATUS
Apparatus
E. m. f . of t w o half-cells is balanced against potentiometer.
The choice of apparatus for the electrometric titration of boric acid should be determined by the instruments available, the number of analyses to be made, and the frequency of use. The requirements for satisfactory operation are:
electrode pairs have been found satisfactory: quinhydrone and 0.1 iV calomel; quinhydrone and saturated calomel; glass and saturated calomel. There are, no doubt, other combinations that might be used but their suitability should be considered, taking as a criterion the above statements on sensitivity and stability.
Sensitivity. At the end point, in an unbuffered solution, a single drop of 0,0231 N sodium hydroxide should deflect the galvanometer 5 to 10 scale divisions. The galvanometer used has a sensitivity of 0.025 microampere per scale division. Stability. The end point should be stable with little if any more drift than occurs in the potentiometric measurement of pH. In this connection, it has been observed that traces of iron in the quinhydrone cause a slow drift t o the acid side. Rapid equilibrium. Some of the high-resistance glass electrodes would probably be unsatisfactory because of the slowness with which equilibrium is reached.
Procedure PREPARATION OF SAMPLE.Remove all foreign matter from the green plant material but avoid excessive washing. Dry a t 70” C., grind, dry again to constant weight, and store in tightly stoppered bottles. If i t is desired to express the results on the green weight basis, weigh the material before and after drying. Weigh an aliquot part of the dry sample and transfer to a glazed paper. From 5 t o 25 grams of the material should be taken, depending on the boron content; there should be not more than the equivalent of 2 mg. of elemental boron i n the portion ti-
Three sets of apparatus are described below, any one of which will give satisfactory results. The first requires a minimum of equipment. The operation 34 l
342
INDUSTRIAL AND ENGINEERING CHEMISTRY
VOL. 12. NO. 6
trated in Table 111. (2) I n the plant parts that normally TABLE I. REPRODUCIBILITY OF IGXITION-ELECTROMETRIC accumulate boron, the concentration is a function of maturity. TITR.4TIOS M E T H O D Leaf samples from Persian walnuts were collected at intervals %-eight of 4 weeks during the growing season. The boron determinaSample Repliof Boron KO. cation Plant Species Sample Found Notes tions, reported in Table IV, show the relationship between Grams 4 f g /kg." maturity and boron concentration. L-701 1 Strawberry fruit, 10 19 It is the practice of this laboratory to collect alfalfa a t the 2 dry 10 18 3 15 18 early flower stage, citrus leaves when 9 to 15 months old, de4 15 19 ciduous leaves at the time of maturity of the fruit, and grasses 5 25 18 6 25 19 a t the time of floT.ver. P-430 1 Alfalfa, field grown 15 10 Characreristic 2 25 10 boron deticiency GLASSWARE.Boron-free glassware is not essential for this P-424 1 Alfalfa, field grom-n 10 60 Normal > i_ n_ HO determination, provided that heating in alkaline solution is P-519 1 Alfalfa, greenhouse 4 i90 Dwarfing due t o not prolonged. Blank determinations completed in Pyrex L culture 1 191 boron toxicity glassware were not significantly different from those in Kaa M g . per kg. of dry material valier brand boron-free glassware. The time of heating on TABLE 11. RECOVERY OF BORON the steam bath was about 0.75 hour. Table V reports the re(Added as crystalline boric acid, to, I 0 grams of plant material prior t o sults of several experiments with Pyrex glassware. ignition) Referring to the second entry in Table V which, in length Sample RepliInitial Boron Boron Boron Recovered Added Found So. cation Content of time of heating, is similar to the procedure suggested, i t is MQ. Mg./kQ.a M g . Xg. Ny. 5; probable that most of the boron found in the bolution came L-701 1 18 0.18 1.10 1.28 1.10 100 from the reagents, as a similar experiment with a Kavalier 2 1s 0.1s 1.10 1.28 1 10 100 3 18 0.18 1.50 1.70 1.52 101 brand boron-free beaker yielded essentially the same quantity P-424 1 60 0.60 U.94 1 54 0.94 100 of boron. a From Table I. IKTERFERIKG SUBSTAXCES.Phosphate reacts with mannitol, though not quantitatively, and must be removed prior to titration. This can be done with calcium, barium, or lead in trated. For each gram of the sample, add 0.1 gram of calcium alkaline solution, but if either sodium hydroxide or sodium oxide and mix well on the paper. Transfer to a porcelain casserole or platinum dish, ignite as completely as possible in a muffle carbonate is used to make the solution alkaline, a fine, gelatiat low red heat, cool, and moisten with water. Cover with a nous precipitate results that adsorbs measurable quantities watch glass, introduce 15 to 20 ml. 6 N hydrochloric acid, which of boron. I n this procedure, lead nitrate and sodium bicarshould make the solution strongly acid, and heat on a steam bath bonate are suggested. It is assumed that the excess lead is for 30 minutes. Cool, transfer to a beaker, and add 1 N lead nitrate solution, precipitated as the normal carbonate rather than as a basic 1 ml. for each gram of plant material in the sample, to remove carbonate. I n any event, the occlusion of boron by the phosphate from the solution. Add sodium bicarbonate, 1 gram precipitate is negligible, as reference to Table I1 will show. at a time, until a precipitate forms. Heat on a steam bath and Germanium, as germanic acid, reacts with mannitol to continue adding sodium bicarbonate until neutral to bromothymol blue. The color should be blue-green, and the pH approxiliberate hydrogen ion, but i t is not a normal constituent of mately 7.0. Make to 250 ml., mix thoroughly, filter through a plant ash in excess of spectroscopic quantities. It has been dry filter paper, and take an aliquot for electrometric titration. shown that small quantities are removed from solution by To expel carbon dioxide prior to titration, make acid with 6 N the lead nitrate precipitation. hydrochloric acid and heat to boiling. Stir cautiously at first, then vigorously. Remove from the iire, make alkaline with 0.5 N Iron, aluminum, and probably most of the weak anions, sodium hydroxide, acidify with 2 N hydrochloric acid and add 5 to that would tend to buffer the solution, are removed by this 10 drops in excess. Make to about 300 ml., again heat t o boiling, process. and stir. Avoid boiling more than a few minutes. Cool. DeOther substances, present in plant ash, appear to be without termine boron by electrometric titration as directed below. ELECTROMETRIC TITRATION. With the galvanometer switch effect. open, introduce the electrodes and stirrer into the solution. Start REAGEhTS. The standard 0.023 N sodium hydroxide can the stirrer and add carbonate-free 0.5 N sodium hydroxide to be stored in paraffin-lined bottles or in copper tanks, the seams ap roximate neutrality as shown by the bromothymol blue. A& about 0.2 gram of quinhydrone. Close the switch. The galvanometer should indicate approximate balance. If it swings to the right, excess alkali is indicated, and if to the left, excess IN MAMMOTH RUSSIAN SUNFLOWERS TABLE 111. BORON acid. Adjust with either 0.0231 N sodium hydroxide or 0.02 (Grown in a nutrient solution containing 2.5 p. p. m. of boron) N hydrochloric acid. The galvanometer should be steady, showLaboratory KO. Plant P a r t Boron ing at most only a slow drift. This is the initial point of titra.Mg./iZg. tion. Add 5 * 0.1 grams of mannitol. If boric acid is present, the indicator will change to the acid color and the galvanometer P-487 Leaves, petioles 597 P-488 Stems 27 will swing to the left. Add standard sodium hydroxide until P-489 Roots 40 balance is again attained. This is the end point. Note the number of milliliters of sodium hydroxide required WALNUT LEAVES TABLE IV. BORONIN PERSIAX after adding the mannitol at the initial point of titration. The (Collected a t intervals of four weeks during growing season) buret should be of such accuracy that the volume of alkali can be estimated to 0.01 ml. A blank is determined by substituting some Laboratory No. D a t e of Collection Boron boron-free organic material, such as ashless flter paper, for the Mg./kg. plant sample and proceedmg as indicated above. From the 166 July 22, 1929 L-199 gross volume this blank is subtracted. The blank used in the 257 August 19, 1929 L-256 301 September 17, 1929 work represented below varied from 0.14 to 0.17 ml. of standard L-306 346 October 14, 1929 L-312 0.0231 N sodium hydroxide. The equivalency of the standard sodium hydroxide with respect to boron is established by titratBORON DISSOLVED FROM 400-ML. PYREX BEAKERS TABLE V. ing an aliquot of the boric acid solution. Report the results as BY HOTALKALINE SOLUTIONS milligrams of boron per kilogram of dry plant material. Dissolved Boron
Discussion
COMPARABLE SAMPLES. I n collecting plant material for boron studies, two facts should be recognized: (1) Boron is not uniformly distributed throughout the plant, as is illus-
Solution
Treatment
hfl. 260 250 250
Water only Saturated NaHC03 Saturated NaHCOa
Steambath Steam bath Steam bath
Hours
MQ.
0.5 0.5 24
0 0.02 0.69
JUNE 15, 1940
ANALYTICAL EDITION
of which should be brazed rather than soldered. The solution should be protected from carbon dioxide and ammonia by an absorption train, preferably a n acid and a n alkali in solution, as dry reagents tend to absorb moisture. Loss OF BORONDURING IGNITIOK. The addition of an alkali before ignition appears to be necessary in the case of certain samples, as, for instance, the dry strawberry fruit (Sample L-701). In a n experiment similar to that reported in Table 11,but in the absence of calcium oxide, the recovery of added boron was only 84 per cent. Dodd (3)reported that boron is lost, even in the presence of alkali, when fatty substances are ignited, and suggested that the fat be extracted with ether or benzene prior to ignition. This procedure was tried in the case of sunflower seeds with results that tend to support Dodd’s findings, although the indicated loss mas small.
343
Suinmarv The electrometric titration method has been adapted to the determination of boron in plant material. It is particularly suited to low concentrations found in boron deficiency studies -e. g., in the range below 50 mg. of boron per kilogram of dry material.
Literature Cited (1) Am. Pub. Health Assoc., “Standard Methods for Examination of Water and Sewage”, 8th ed., pp. 30-3, 1936. (2) Brandenburg, E., Phytopath. Z., 3, 499-517 (1!331). (3) Dodd, A. S.,Analyst, 54, 716-25 (1929). ( 4 ) Kelley, W. P., and Brown, 9. hl., Hilgardia, 3, 445-58 (1928). ( 5 ) McMurtrey, J. E., Jr., J . Am. SOC.Agron., 27, 271-3 (1935). (6) Warington, K . , Ann. Botany. 37, 630-70 (1923). (7) Wilcox, L. V., IND. ENG.CHEM.,-4nal. Ed., 2, 358-61 (1930). (8) Ibid., 4, 38-9 (1932).
Rapid Determination of Copper By a-Benzoin Oxime in Ferromolybdenum, Calcium Molybdate, Etc. LOUIS SILVERMAN, 5559 Hobart St., Pittsburgh, Penna.
F
E I G L (3) introduced a-benzoin oxime, Cd%.C(SOH).CHOH.C&, as a specific reagent for copper in ammoniacal solution. He suggested that the compound formed v-as one of new structure in which the principal valences of copper are connected through the oxygens of the oxime and hydroxyl groups, respectively, and the coordinate valences are attached to the phenyl groups. Azzalin (I), however, found (1) that for direct weighing procedures excess reagent could not easily be washed away; (2) that for ignition procedures only slow and careful ignition would remove all carbon; and (3) that in the presence of other elements the precipitate is contaminated with these elements. Obviously, then, the final determination of the copper should be made either by electrodeposition or by the short iodide titration (2, 6, 8). The rapid determination of copper in ferromolybdenum or in molybdenum steels is thus accomplished.
Proposed Procedure PREPARATION OF SAYPLE. For analyses where silicon is also to be determined ( 6 ) ,weigh 1 gram! place in a tall 300-cc. beaker, and treat with 20 cc. of (1 2) nitric acid. After rapid action ceases, heat the beaker until solution of the sample is about complete, cool somewhat, and cautiously add 10 cc. of concentrated sulfuric acid. Place the beaker in a bumping beaker, and heat to heavy fumes of sulfuric acid. Cool, add 75 cc. of (1 4) hydrochloric acid, heat until soluble salts are dissolved, and filter at once. Wash alternately with cold (1 4) hydrochloric acid and water six times, then with hot solutions to remove iron completely. The filtrate is used for co per. ’Ct’here silicon is not required, ogtain complete solution by the above nitric acid mixture with the aid of some hydrofluoric acid. After cooling, and diluting with water, add 10 cc. of hydrochloric acid. For molybdenum steels, the nitric acid solution method may be folloR-ed by fuming with perchloric acid. After cooling, add water and hydrochloric acid. For calcium molybdate, use 25 cc. of ( 2 1) hydrochloric acid. For ferrotungsten, use 10 cc. of (1 4) nitric acid, 5 cc. of hydrochloric acid, then hydrofluoric acid dropwise t o complete solution. PRECIPITATION OF COPPER. Dilute the prepared solutions to about 150 t o 200 cc., and add a solution of 15 grams of Rochelle salt dissolved in 15 cc. of water. Add a strong solution of sodium hydroxide until the test solution is alkaline t o Congo red paper, but still acid to litmus paper. If necessary, allow t o cool to room temperature. Add ammonium hydroxide till the solution turns
+
+
+
+
*
+
blue (copper above 0.5 per cent), or until distinctly alkaline to litmus, and add about 5 cc. in excess. Add slowly, with stirring, a 2 per cent a-benzoin oxime solution (10 cc. for 0 to 0.5 per cent, 15 cc. for 0.5 to 1.5 per cent copper). Let stand about 15 minutes, filter the green precipitate on an 11-cm. No. 40 Whatman paper, and wash with warm (1 99) ammonium hydroxide solution. DETERMISATIOS OF COPPER. Return the paper and contents to the beaker and add 15 cc. of nitric acid (specific gravity 1.4) and 10 cc. of perchloric acid (60 or 70 per cent grade). Mix, heat until the nitric acid has been driven out and the perchloric acid condenses on the walls of the beaker, cool, dilute with water, and boil out the chlorine. [If desired, the short iodide titration (2, 5 , 8 ) for copper may be used instead, starting.at this point.] Add ammonium hydroxide until the solution is blue, discharge the color with nitric acid, add 4 cc. of (1 1) sulfuric acid, and plate the copper at 3 volts and 0.5 ampere.
+
+
Discussion The reasons for this particular procedure should be noted. After solution of the sample and addition of the sodium potassium tartrate, the sodium hydroxide is added until the solution is alkaline to Congo red paper but acid to litmus (pH 4 to 6). If enough caustic is added to turn litmus blue, sample KO.26789 (1.00 per cent copper) returns only a portion of the copper (0.60 per cent) even after 3 hours’ standing. Evidently the alkaline tartrate has partially removed cupric ion from solution. Again, precipitation should take place at room temperature. K h e n precipitation was made hot, the results mere low and inconsistent. It is conceivable that some of the oxime was hydrolyzed into ketone and hydroxylamine, which might reduce some cupric ions to the cuprous stage and be held in solution by the ammonia and ammonium salts. Feigl stated that there should be no ammonium salts present. I n the final determination of the copper either by electrodeposition or by titration, impurities such as excess reagent, molybdenum, nickel, manganese, iron, etc., do not interfere. I n many cases the colorimetric ammonia method could be used. The usual methods for separation of copper from ferromolybdenum require separation from the molybdenum by use of caustic (or fusion), then separation of the copper from the iron-operations which are slow and undesirable. The pro-
